Skip to main content
NeuroBrix

Enterprise & Services

Local AI on your own hardware

NeuroBrix runs AI models for text, images, audio, video and multimodal tasks on machines you control. For organisations that want it inside their own infrastructure, consulting, deployment, engineering and support are distributed commercially by WizWorks OÜ, the distributor of NeuroBrix.

Who it is for

Organisations that want to run AI locally: to keep their data on their own premises, to use hardware they own or choose, and to depend on no external AI service for their models.

What we offer

  • AI strategy & consulting

    Choose the models that fit your use cases and the hardware you have, and plan a local deployment step by step.

  • On-premise deployment

    NeuroBrix installed on your own servers, serving models through an OpenAI-compatible API that your existing tools can call.

  • Custom engineering

    Bring a model you need into the .nbx format, tune the runtime for your hardware, or integrate it with your systems.

  • Support & training

    Help for the teams who run NeuroBrix day to day: installation, model management and serving, with training on request.

How it works

NeuroBrix stays open source

The engine is created by Hocine Benkelaya and released under the Apache License 2.0. The engine, its documentation and the model hub stay freely available, with or without a services agreement.

Distributed commercially by WizWorks OÜ

WizWorks OÜ, the distributor of NeuroBrix, is an AI company registered in Tallinn, Estonia, with its lab in Alicante, Spain. It runs its own GPU infrastructure and builds AI products on it, from the hardware up.

wizworks.io

Tell us what you need

Your request goes to NeuroBrix and to WizWorks OÜ at the same time.